SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

We have investigated the physical, biochemical, and cellular properties of an autologous leukocyte and platelet-rich fibrin patch. This was generated in an automated device from a sample of a patient's blood at the point of care. Using microscopy, cell counting, enzyme-linked immunosorbent assay, antibody arrays, and cell culture assays, we show that the patch is a three-layered membrane comprising a fibrin sheet, a layer of platelets, and a layer of leukocytes. Mean recovery of platelets from the donated blood was 98% (±95%CI 0.8%). Mean levels of platelet-derived growth factor AB, human transforming growth factor beta 1, and vascular endothelial growth factor extracted from the patch were determined as 127 ng (±95% CI 20), 92 ng (±95%CI 17), and 1.35 ng (±95%CI 0.37), respectively. We showed a continued release of PDGF-AB over several days, the rate of which was increased by the addition of chronic wound fluid. By comparison with traditional platelet-rich plasma, differences in immune components were found. The relevance of these findings was assessed by showing a mitogenic and migratory effect on cultured human dermal fibroblasts. Further, we showed that fibrocytes, a cell type important for acute wound healing, could be grown from the patch. The relevance of these findings in relation to the use of the patch for treating recalcitrant wounds is discussed.

Chronic wounds remain a challenge for patients as well as health providers. Treating chronic wounds with autologous growth factors has shown promising but diverse results. The main ways of providing growth factors for the treatment of wounds has been by the use of recombinant PDGF-BB (Regranex gel, Systagenix, Gatwick, United Kingdom) or the generation of platelet-rich preparations, either based on platelet concentrates, platelet-rich plasma (PRP), or recently, platelet-rich fibrin (PRF). PRP-based treatments are performed by activating a PRP suspension with thrombin, typically either bovine (Autologel System, Cytomedix, Gaithersburg, MD) or autologous (Biomet, Warsaw, IN), and/or calcium ions, thereby generating a platelet gel incorporating the activated platelets.[1] Several ways of generating platelet-based treatments for wound treatment have been developed based on either allogenic[2] or autologous material. The latter ranges from simple laboratory instructions using standard equipment[3] to complex systems with dedicated reagents, hardware, and utensils (Vivostat [Allerød, Denmark], GPS [Biomet], and SmartPrep [Harvest Technologies, Plymouth, MA]). Studies comparing blood cell, particularly platelet, recoveries, and growth factor release from the resulting preparations generally show that there is a high diversity in composition between different preparations and within preparations produced by the same device when determined on different donors.4,5

Clinical evidence for the effectiveness of platelet-rich products is limited, mainly due to small uncontrolled studies. However, a recent meta-analysis of published studies found that PRP favors the healing of diabetic foot ulcers.[6] PRF could have several advantages over PRP and other platelet-rich products due to its high fibrin content; these include greater mechanical strength, extended release of growth factors, and the protection of growth factor against proteolysis.7,8 Studies have shown that growth factors derived from platelets interact with the fibrin (vascular endothelial growth factor [VEGF],[9] basic fibroblast growth factor,[10] fibroblast growth factor 2 [FGF-2][11]). Furthermore, studies show that these interactions could be the basis for protection from proteolytic degradation7,12 and growth factor potentiation.[13] Although existing PRFs are similar to PRP in terms of growth factor contents, only a few clinical studies have been performed in nonhealing wounds.[14]

In order to be clinically relevant, the treatment should have an effect on wound healing and the processes involved, including the formation of granulation tissue. Granulation tissue is mainly formed by the fibroblast and endothelial cells resident in the wound and its surrounding tissue. Additionally, studies have shown that a subpopulation of circulating leukocytes, the fibrocytes, is involved in wound healing.[15] In acute wound healing, these cells migrate to the wound[16] and are involved in several processes important for proper healing.[17] Fibrocytes are described by their hematopoietic origin combined with the synthesis of extracellular matrix proteins (e.g., collagen).18,19

In this study, we characterize a new platelet and leukocyte-rich fibrin patch produced by a two-spin process in a closed sterile system. In contrast to the PRF produced by others (e.g., Dohan et al.[20]), the patch described displays a three-layered structure as well as a high content of platelets and leukocytes. We investigate the structure and ultrastructure of the patch as well as the release of selected growth factors. To relate results to traditional PRP, we compare released growth factors, cytokines, and chemokines by antibody arrays as well as enzyme-linked immunosorbent assay (ELISA) of selected factors. Further, we determine the effect of patch-derived factors on normal human dermal fibroblasts (NHDF) cells and the release of autologous fibrocytes. Finally, we discuss the relevance of these characteristics to its proposed use for the treatment of difficult to heal chronic wounds.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Preparation of an autologous leukocyte and platelet-rich fibrin patch

Blood was collected by venepuncture from an eligible donor. Blood (16 mL) was drawn directly into a Leucopatch™ Device (Reapplix, Birkerød, Denmark) without anticoagulant using a winged blood sampling set (Terumo Quick fit, cat. no. MN-SV21Q, Terumo, Leuven, Belgium) fitted with the Leucopatch needle holder (Reapplix). The device was transferred directly to the provided centrifuge holder and centrifuged for 8 minutes at 3,000×g in an Eppendorf centrifuge (cat. no. 5702, Eppendorf, Hamburg, Germany). After an additional 10 minutes incubation, the device was removed from the provided centrifuge holder and spun for a further 2 minutes at 3,000×g to lift the filter system included in the device and thereby compacting the polymerized fibrin into a platelet and leukocyte containing patch. After removing the device lid the patch was transferred for further analysis.

Characteristics of blood used for manufacture

A total of 10 healthy volunteers were used as donors for the studies performed. Informed consent was obtained from each donor and approval for the study was received from the local ethics committee for research in human subjects in accordance with the Declaration of Helsinki. Vacuum containing Leucopatch devices were filled by venepuncture as described by the manufacturer (Reapplix), and additionally a 4 mL control sample was drawn into an EDTA tube (Terumo, Venosafe, cat. no. VF-054SDK) for blood analysis using a Sysmex SF3000 cell counter (Sysmex, Kobe, Japan). Donor and baseline blood characteristics are given in Table 1.

Table 1. Platelet and leukocyte numbers and recoveries
 Donor full blood average (n = 10) (×106 cells pr. mL)Donor full blood range (n = 10) (×106 cells pr. mL)Patch average (n = 10) (×106 cells)Recovery* (%)Recovery 95% confidence interval
  1. *Percentage of full blood platelets and leukocytes retained in the patch (cells in patch/blood volume drawn × full blood cell number × 100%).

Platelets218151–3823528980.8
Neutrophils2.732.05–3.9624.65510
Lymphocytes2.031.17–2.4925.3773
Monocytes0.420.34–0.535.9854

Determination of the physical structure

Patches were transferred directly from the device to either 10% formalin (formaldehyde 4% aqueous solution, buffered, cat no. 9713.1000, VWR, Herlev, Denmark) or 2.5% glutaraldehyde (glutaraldehyde solution, cat no. 49632, Sigma Aldrich, Brøndby, Denmark) in Dulbecco's phosphate-buffered saline (DPBS; cat no. D8537, Sigma Aldrich) for light microscopy and electron microscopy, respectively. For light microscopy, the fixated patches were dehydrated, paraffin embedded, and cut into ultra thin (3–5 μm) sections using a microtome (InLab, Virum, Denmark). This was followed by rehydration and hematoxylin–eosin (H&E) staining. For electron microscopy, patches were dehydrated and mounted on aluminum stubs before coating with approximately 20 nm gold plating to make the surface conducting. For cryo-scanning electron microscope (SEM), the freshly prepared patches were fixated in liquid nitrogen before transferred under vacuum into the SEM. In the SEM chamber, the frozen sample is located on a cold stage. From fixation to investigation of the sample, the temperature was kept lower than −120 °C. A Zeiss Ultra 55 (Zeiss, Oberkochen, Germany) SEM system was used to photograph ultrastructures.

Determination of cellular composition

Three patches were produced for each donor as described previously. In order to obtain baseline cell levels, a sample was taken into a 4 mL EDTA tube (Terumo, Venosafe, cat. no. VF-054SDK) simultaneously. After processing and removal of the generated patches (including all polymerized fibrin), the remaining blood components were mixed thoroughly to obtain a homogenous solution. In order to assure that no cell aggregates or polymerized fibrin were left, the solutions were filtered using a 40 μm filter (cat. 07–40/25, Sefar Inc, Ruschlikon, Switzerland), which after filtration were microscopically inspected for any blood components. Cell numbers in all samples were determined using a Sysmex SF-3000 cell counter (Sysmex). Cell recoveries were calculated as percent of cells removed from the blood within the patch relative to the number of cells in the blood used for its production.

Growth factor extraction

Growth factors were extracted from patches by shaking and three freeze-thaw cycles. The freshly prepared patch was transferred to 4.5 mL phosphate-buffered saline (PBS) 1% bovine serum albumin (BSA, BP1605-100, Fisher Scientific, Waltham, MA) containing protease inhibitor (Complete Mini, cat.no. 11 836 153 001 at 1 tablet per 10 mL PBS 1% BSA as described by Roche, Basel, Switzerland) and vigorously shaken for 30 minutes prior to freezing at −80 °C. Before growth factor detection, the samples were thawed and refrozen twice for a total freeze-thaw cycle number of three. Samples were centrifuged for 15 minutes at 3,700×g and the supernatant used for growth factor detection. Protease inhibitors were included to exclude any proteolytic component in the freeze-thaw-based growth factor release.

Growth factor release over time

Four patches from each of two donors were generated as described previously.

After generation, two patches from each donor were transferred to 4.5 mL of 1% BSA in PBS (PBSBSA) in a 15 mL tube and incubated at room temperature with end-over-end mixing. In parallel, two other patches from each donor were transferred to 2 mL of 2% chronic wound fluid in PBS (PBSCWF) in a 15 mL tube and incubated as above. Conductivity and pH were determined for both dilutions.

Chronic wound fluid (CWF) was obtained from a patient with a chronic diabetic foot ulcer (>5 years duration) after informed consent and approved by the Regional Research Ethics Committee of Copenhagen (CVK approval 19632) in accordance with the declaration of Helsinki. A Biatain (Coloplast, Humlebaek, Denmark) foam bandage, removed after 24 hours, was centrifuged at 3,700×g—the exudate fluid was sterile filtered (0.22 μm) and kept at −80 °C until dilution in PBS.

After 30 minutes, 2 hours, 4 hours, 24 hours, 48 hours, 72 hours, 120 hours, and 168 hours of incubation, the patches were transferred to new tubes containing either 4.5 mL PBSBSA or 2 mL 2% PBSCWF. Immediately after transfer, the used buffer were frozen at −80 °C until growth factor determination.

Growth factor concentration (ELISA)

The levels of platelet derived growth factor AB (PDGF-AB), human transforming growth factor beta 1 (TGF-β), interleukin 8 (IL-8), and vascular endothelial growth factor (VEGF) were determined by commercially available ELISA assays (DuoSet ELISA Development System, human PDGF-AB cat. No. DY222, human TGF-β1 cat. No. DY240, human IL-8 cat. No. DY208 and human VEGF cat. No. DY293B, R&D systems, Abingdon, United Kingdom) as described by the manufacturer. For TGF-β, this included a sample activation step adding 20 μL of 1 M HCl to 40 μL sample for 10 minutes at room temperature (RT) followed by neutralization with 1.2 N NaOH/0.5 M HEPES followed by dilution in Reagent diluent (Cat # DY997, R&D Systems) as described by the manufacturer.

Growth factor amounts released per patch were calculated and results given as mean ng/patch based on three individual devices on each of the 10 donors are given. For comparison purposes, results are also given in pg/106 platelets in the starting blood samples and pg/106 platelets in the processed product. To be able to relate protein levels at the treatment level, ng/cm2 are calculated; for comparisons with PRP treatments, the application of a 2 mm PRP layer is hypothesized.

PRP preparation

PRP was generated as described previously.[7] Briefly, blood was drawn into 3.5 mL ACD tubes (Vacuette, cat. no. 454332, Greiner bio-one, Frickenhausen, Germany) and spun for 20 minutes at 135×g at 4 °C without deceleration. The platelet-rich supernatant (approximately 1.5 mL) was transferred to a new tube and the platelet concentration determined using a Sysmex SF-3000 cell counter (Sysmex). To standardize platelet concentrations, platelets were pelleted at 2,000×g for 10 minutes and a calculated amount of plasma was removed to obtain a final concentration of 109 plt/mL.

Standardized PRP and patches from the same donors were used for comparison studies. PRP or patch (in 4 mL PBS) was frozen at −80 °C followed by thawing at 37 °C—this procedure was repeated twice to obtain a total of three freeze-thaw cycles. Samples were centrifuged at 10,000×g for 15 minutes. Supernatants were used for ELISA and antibody arrays as described.

Detection of cytokine, chemokine, and growth factors by antibody arrays

PRP and patch-derived extracts were tested for cytokine, chemokine, and growth factor content using antibody arrays. The commercially available human cytokine array panel is able to detect the relative levels of 36 cytokines (cat. No. ARY0005. R&D Systems) while the human angiogenesis array detects 55 angiogenesis-related proteins (cat. No. ARY007. R&D Systems); due to six overlaps, a total of 85 proteins could be detected. The antibody arrays were used as described by the manufacturer. Samples from PRP and patch extracts from three donors were pooled. To be able to compare PRP and patch samples, they were standardized by dilution in the provided assay buffer to obtain identical PDGF-AB levels as determined by ELISA. Samples were preincubated with an antibody cocktail for 1 hour and then incubated at 5 °C for 16 hours with the provided nitrocellulose-based antibody arrays as described by the manufacturer. Final detection was performed by addition of chemi-luminescence substrate (Immobilon Western, cat. No. WBKL S0050, Millipore, Copenhagen, Denmark) and analyzed using a luminescent image analyzer system, Alpha Innotech FluorChem 8000 (San Leandro, CA). Mean pixel density was determined for each array spot and negative control values subtracted. To assure relevant signals, only values 3× above those of negative control values were used for further analysis; a twofold increase/decrease between samples was used as a minimum to screen for differences.

Effect on fibroblast growth

The effect of the patches on primary human fibroblast cells was determined as described previously,[7] although fitted to an insert based cell culture system.

Briefly, primary adult NHDF-As (CC-2511, Lonza, Basel, Switzerland), derived from a 37-year-old Caucasian female, were purchased from Lonza and expanded in fibroblast growth medium (FGM) with 10% fetal calf serum (FCS; Invitrogen, Carlsbad, CA) in 75 cm2 culture flasks (Nunclon®, Nunc, Roskilde, Denmark) at 37 °C in humidified 5% CO2/air. FGM is composed of fibroblast basal medium (FBM; CC-3131, Lonza) containing gentamicin and amphotericin-B, and supplemented with growth factors (SingleQuotss®, CC-4134, Lonza). Cells subcultured four times according to Lonza's recommendation were used for the assay. In the assay cells were grown in FBM added 2% FCS (FBM 2% FCS) in Nunc 96 well 0.2 μm Anapore membrane inserts (cat #136730, Nunc) submerged in wells added FBM 2% FCS and a 2 mm biopsy of patch. Inserts without NHDFs (patch control) or wells without patch added (FBM 2% FCS) were used as controls. At the indicated time points, inserts were removed and the cell numbers in the inserts were determined by adenosine triphosphate (ATP) detection using a luminescence-based assay (Vialight Plus kit, cat LT07-221, Lonza). Cell numbers were calculated using ATP and NHDF cell standard curves. Samples were run in quadruplicates and results were shown as averages with 95% confidence intervals.

Fibrocytes

Fibrocytes were grown from patches essentially as described by Quan and Bucala.[21] Briefly, wells on an 8-well chamber slide (LAB-TEK II, NUNC 154534, Nunc) were coated by adding 200 μL of sterile-filtered 10 μg/mL fibronectin (F1141, Sigma-Aldrich) in DPBS per well.

Coating was performed at 37 °C for 1 hour. Chambers were emptied and Dulbecco's modified Eagle's medium (DMEM, E15-843, PAA Laboratories, Pasching, Austria) with 20% FCS was added at 350 μL/well. Patches were generated as described previously and transferred to a sterile Petri dish upside down (leukocyte surface up); 4 mm biopsies were made and one transferred to each well, leading to approx. 4.5 × 105 leukocytes/cm2. Chamber slides were incubated at 37 °C with 5% CO2. After 48 hours, patches were removed and the DMEM media changed. DMEM 20% FCS was changed again after another 6 days. After 14 days, the attached cells were fixed in −20 °C methanol for 10 minutes and washed three times with DPBS for a total of 10 minutes. Slides were dried and kept dry at 5 °C until immune labeling for collagen I, macrophage marker CD115, and the fibrocyte-associated CD45RO marker.[19]

Immunochemical detection of CD115 and CD45RO was performed as a double labeling on the same slide, and detection of collagen I was performed separately on another slide. After blocking of endogenous biotin and peroxidase activity, the primary antibodies against CD115 (rat IgG2b monoclonal antibody) (Santa Cruz Biotechnology, Santa Cruz, CA) and CD45RO (mouse IgG2a monoclonal antibody) (BD Biosciences Pharmingen, San Diego, CA) were mixed in dilutions of 1:50 and 1:100, respectively, in 1% horse serum + 0.1% Triton X-100 in PBS and incubated overnight at 4 °C. After wash in PBS, a mix of secondary antibodies comprising a 1:200 dilution of biotin-labeled goat-anti-mouse IgG2a (Southern Biotech, Birmingham, AL) and a 1:1,000 dilution of peroxidase conjugated goat-anti-rat IgG (Nordic BioSite, Copenhagen, Denmark) was added to the slide. The CD45RO signal was generated by adding Cy5-labeled streptavidin (Invitrogen, Taastrup, Denmark), and the CD115 signal was established by adding the fluorescent peroxidase substrate Cy3-TSA (Perkin Elmer Life Science, Waltham, MA).

Collagen I immunochemical detection was conducted by using a mouse monoclonal IgG2a Collagen Type I antibody (MP Biomedicals, Illkirch, France). After incubation in a 1:100 dilution of this antibody in 1% horse serum + 0.1% Triton X-100 in PBS overnight at 4 °C, the collagen type 1 signal was generated using a FITC-labeled secondary anti-mouse IgG antibody (Sigma-Aldrich).

Images of the labeled cells were acquired using an Axioimager Z1 epifluorescent microscope equipped with a 20× Plan-Apochromat objective (Zeiss).

Statistical methods

The significance of differences between extracted proteins levels, PDGF-AB release profiles and cellular growth was assessed at each time point using Student's t-test (two tailed, equal variance).

Correlation and linear regression analysis was performed using Prism software (GraphPad Prism version 3.00 for Windows, GraphPad Software, San Diego, CA).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Macroscopic and microscopic structure

The autologous leukocyte and platelet rich patch as produced by the Leucopatch device (Figure 1A), is circular with a diameter of 2.5 cm (area of 4.9 cm2), is approximately 1–2 mm thick, and has a gelatinous appearance (Figure 1B). The patch is generated by a two-step procedure including an initial hard spin that separates full blood into its components and leads to complete coagulation of the separated plasma above the leukocyte and platelet containing buffy coat. This is followed by a second hard spin that due to the construction of the device compacts the polymerized fibrin to form a patch. Whereas the area of the generated patch is constant, the thickness and thereby volume of the patch can vary due to variations in fibrin and cell concentration of the donor. Upon histology preparation and H&E staining, a three-layered structure appears (Figure 1C). The top layer is composed of the compacted fibrin (Figure 2A–C). The center of the patch is composed of a layer of highly concentrated platelets suspended in fibrin as seen by the characteristic pink H&E staining of platelets (Figure 1D) and platelet morphology in SEM (Figure 2D–F). The platelets appear partly activated by the processing as both rounded and stellate cell morphologies are seen by SEM. The bottom layer is composed of leukocytes embedded in a fibrin structure (Figures 1E and 2G–I).

figure

Figure 1. Appearance of the generated patch. (A) The patch being removed from the Leucopatch device; (B) the circular patch shown cell side up; (C) H&E stained histology section showing the three layer—fibrin (top), platelets (light pink, middle), and leukocytes (bottom layer) (25×); (D) platelet layer (200×); (E) leukocyte close-up (630×).

Download figure to PowerPoint

figure

Figure 2. Ultrastructure of patch. (A) Upper fibrin surface by SEM; (B and C) fracture surface of fibrin layer—generated by Cryo-SEM; (D and E) fracture surface of platelet layer; (F) platelets by EM; (G) leukocyte cell side surface by SEM; (H and I) fracture surface of leukocyte layer (Cryo-SEM).

Download figure to PowerPoint

Cellular composition

Studies of the cellular components show that the majority of platelets and leukocytes are transferred to the patch during preparation. Studies on 10 donors (mean age 43.4 years [range 29–58], 4 females), with normal donor cell range (Table 1) representing the A (n = 3), B (n = 1), and O (n = 5) blood types, show that 98% of platelets are transferred to the formed patch while the majority of the leukocytes are also retained. By automated cell counting, the neutrophils, lymphocytes, and monocytes retained in the patch were determined to present 55, 77, and 85% of the starting material, respectively (Table 1). All three types of leukocytes appear intact in both H&E stains and on SEM photos (Figures 1D and 2D–F) and do not seem to be harmed by the patch production process. The existence of intact cells is confirmed by the ability of a subpopulation of these cells to transdifferentiate into fibrocytes upon culturing (Figure 3).

figure

Figure 3. Fibrocytes grown from the patch. (A) Phase contrast; (B) immunofluorescence CD45RO (yellow), CD115 (red), DAPI nuclear stain (blue); (C) collagen 1 (green) DAPI nuclear stain (blue) (200×).

Download figure to PowerPoint

By transferring the cell-containing patch directly to cell growth media, we were able to grow CD45RO positive (Figure 3B) and collagen-producing cells (see Figure 3C) in 14 days. This confirmed the presence of a hematopoietic cell type (CD45RO) that produces collagen and is not a macrophage (CD115 negative, Figure 3B), features defining the fibrocyte cell.[19]

Growth factor concentration

The amount of PDGF-AB, TGF-β, and VEGF from the patches of 10 donors was determined by ELISA. As the volume of the patches can vary depending on plasma contents (between 0.5 and 1 mL), the growth factor contents are given as total amount per patch and amount per area, as well as pictogram per million platelets in the patch. As seen by the ranges given in Table 2, notable variations were found in the growth factor levels between donors. Correlating the growth factor levels of individual donors with their respective full blood platelet numbers shows that this variation can be partly ascribed to the variability in donor blood platelet counts (Figure 4). Whereas the percentage of blood cells recovered is quite consistent from donor to donor (Table 1), the absolute number of cells transferred to the patch is highly dependent on the cell levels of the donor. Consequently, there was a significant correlation between PDGF-AB and full blood platelet numbers; a similar significant correlation was found for VEGF while TGF-β correlation did not reach significance (Figure 4).

figure

Figure 4. Correlation of growth factor concentrations and platelet concentrations in full blood. The linear line of best fit and 95% confidence intervals are shown. Pearson correlation coefficients were r = 0.7244, r = 0.6565, and r = 0.4113, and p = 0.0178, p = 0.0392, and p = 0.238 for PDGF-AB, VEGF, and TGF-β, respectively. PDGF-AB, platelet-derived growth factor AB.

Download figure to PowerPoint

Table 2. Growth factor levels
 Total amount of growth factor released (ng, n = 10)RangeGrowth factor released per 106 platelets in starting blood sample (pg)Growth factor per cm2
  1. PDGF-AB, platelet-derived growth factor AB; TGF-β, human transforming growth factor beta 1; VEGF, vascular endothelial growth factor.

PDGF-AB127.390–23537.126.0
TGF-β91.542–16327.2718.7
VEGF1.350.49–3.650.3810.28

Release of PDGF-AB over time

As previous studies with PRP and PRF have indicated that growth factor release profiles are dependent on the method of platelet activation[22] as well as milieu,[7] we determined the release of PDGF-AB over time. The release of PDGF-AB from patches was determined under two conditions: incubation in PBS 1% BSA or incubation with PBS 2% CWF. Analysis of the two extraction solutions in regard to pH and conductivity showed no differences (pH 7.04 vs. 6.96, conductivity 29.7 mV vs. 26.1 mV for PBSBSA and PBS 2% CWF, respectively), indicating that the differences found in the effect on release were due to other factors. As seen in Figure 5, the release profile of PDGF-AB from the patch is highly dependent on the milieu, whereas the levels released over time are similar; the rate of release is increased in the presence of CWF. Similarly, the physical degradation (solubilization) of the patches was complete after 72 hours of incubation with CWF. In contrast, incubation with PBS 1% BSA did not lead to significant degradation (data not shown). A significantly increased PDGF-AB level was seen after 24 and 48 hours of incubation (t-test, p < 0.01).

figure

Figure 5. Release of PDGF-AB over time. Patches were generated from two donors (donor 1 squares; donor 2 diamonds) incubated in either PBS 1% BSA (dashed line) or PBS-2% chronic wound fluid (CWF, dotted line). *Significant difference between BSA and CWF, p < 0.05. BSA, bovine serum albumin; PBS,phosphate-buffered saline; PDGF-AB, platelet-derived growth factor AB.

Download figure to PowerPoint

Comparison of PRP and patch

In order to compare the described patch with PRP, patches and PRP standardized to 109 plts/mL were made from five donors, and the PRP did not contain detectable leukocytes (below 104 cells/mL). Levels of PDGF-AB, VEGF, and IL-8 were quantified and compared (Table 3). Absolute levels of all three proteins were found to be significantly higher in the prepared patch compared with PRP (PDGF: p < 0.05, VEGF and IL-8: p < 0.01). Similarly, if calculated as protein per 106 platelets in the starting blood samples, levels of detected PDGF-AB, VEGF, and IL-8 were significantly higher in the patch preparation (p < 0.05). In contrast, if extracts were assessed according to platelet level of the final preparation, PDGF-AB levels per 106 platelets were significantly higher in the PRP (p < 0.001), IL-8 levels were higher in the patch (p < 0.05), as well as VEGF was found to be higher in the patch although not significant (p = 0.12).

Table 3. Comparison patch and platelet rich plasma extracts
  Growth factors released (range)Growth factor released per 106 platelets in starting blood sample (pg) (range)Growth factor released per 106 platelets in final product (pg) (range)Growth factors* released per cm2 (ng)
  1. *Compared with the application of a 2 mm layer of PRP.

  2. IL-8, interleukin 8; PDGF-AB, platelet-derived growth factor AB; PRP, platelet-rich plasma; VEGF, vascular endothelial growth factor.

PDGF-ABPatch171.5 ng (123–257)40.1 (37–48)40.9 (37–48)35
PRP66.2 ng/mL (58–72)31.6 (27–39)66.2 (58–72)13
VEGFPatch2.08 ng (1.16–5.09)0.44 (0.30–0.78)0.45 (0.30–0.78)0.36
PRP0.21 ng/mL (0.018–0.59)0.09 (0.01–0.23)0.209 (0.018–0.59)0.042
IL-8Patch1.78 ng (0.57–3.65)0.41 (0.18–0.93)0.42 (0.18–0.93)0.42
PRP0.0063 ng/mL (0–0.108)0.003 (0–0.005)0.006 (0–0.011)0.0013

To compare PRP and patch-extracted proteins more comprehensively, antibody arrays were used. As array results are nonquantitative, we adjusted the sample dilution to obtain identical PDGF-AB levels. After screening for the presence of 85 different proteins, we found 54 that generated a signal above 3× negative control values. Ten proteins showed above twofold differences between patch and PRP samples, eight were higher in the patch; PTX3 (2.4-fold), CD40L (2.4-fold), HGF (2.6-fold), VEGF (2.6-fold), FGF-2 (2.9-fold), IL-1ra (3.4-fold), IL-16 (4.6-fold), and IL-8 (14-fold). And two were higher in the PRP preparation: leptin (3.6-fold) and MMP-8 (threefold) (see Supporting Information Figure S1).

Fibroblast growth

The growth of adult NHDFs was increased in the presence of the patch compared with cell growth in control media (see Figure 6). A significant (p < 0.01) increase in cell numbers was seen in the presence of patch, compared with untreated control wells, at time points 48, 96, 120, and 144 hours. While the control conditions led to a tripling in the number of fibroblast over 6 days, the addition of patch resulted in an almost 14-fold increase in cell numbers during that period. Performing the insert-based assay without fibroblast cells (Figure 6, patch control) did not lead to the detection of ATP in the wells, confirming the fibroblast specificity of the results obtained.

figure

Figure 6. Fibroblast growth in response to the patch. *Significant difference between control (FBM 2% FCS) and patch treated (FBM 2% FCS + Patch); p < 0.05. FBM, fibroblast basal medium; FCS, fetal calf serum.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Advances in wound care products in the recent decades have mainly been due to the development of advanced materials able to maintain a wound milieu optimal for the intrinsic wound healing. Despite this effort, there is consensus that nonhealing wounds benefit from treatments actively affecting the wound-healing cascade.[23]

We have characterized a leukocyte and platelet rich patch in regard to its physical as well as biochemical and cellular properties. The three-layered compacted nature of the patch (Figures 1 and 2) differs it from traditional preparations based on autologous blood. This is due to the surprising finding that at high centrifugation speed, no red blood cell coagulum is formed; rather, a coagulum binding only platelets and leukocytes of the buffy coat is generated. By compacting this coagulum by mechanical means as in the Leucopatch Device, a patch with a defined structure is formed.

The donors used in this study represent both males and females (six and four, respectively) as well as both A, B, and O blood types. Despite observations that raise the possibility for a greater tendency for blood clot formation in non-O patients, probably due to differences in Von Willebrand factor plasma half-life,[24] we did not see any difference in coagulation between O and non-O donors nor between male and female.

As it was not feasible to count cells that were incorporated onto the patch, we determined cell recovery by counting cells remaining in the device after removal of the patch. We assessed the remains by microscopy and filtration to assure that it was a homogenous solution that was measurable by an automated (Sysmex) cell counter. As the Sysmex detection system uses an impedance-based detection method that has previously shown to flag platelet clumps with adequate sensitivity,[25] we are confident in the method used. We hypothesize that the high platelet recovery rates as well as the absence of red cells in the patch are due to the complete compaction of red blood cells during the initial hard spin, and this lead to a near complete separation of red blood cells, platelets, and plasma. Upon polymerization, fibrin binds the buffy coat but the smoother compacted red cells are not bound. This theory fits well with the recoveries seen in our study as the cells with the highest specific gravity (SG), neutrophils (SG: approx. 1.08), have the lowest recovery (55%); in contrast, the denser lymphocytes (SG: approx. 1.07) and monocytes (SG: approx. 1.06) show higher recovery rates (77% and 85% respectively). The lower density of the platelets (SG: 1.05) leads to a recovery of 98%. That a near complete equilibrium is reached is confirmed by the layered structure of the produced patch having leukocytes below the platelet layer (Figure 1C). Further, the significant correlation between full blood platelet numbers and the extracted PDGF-AB and VEGF factors supports the high recovery rate seen.

There is a high degree of variation between PRP/PRF systems, and because of variable ways to measure the cell recovery and growth factor release, direct comparisons are difficult. Data reviewed by Mazzucco et al. described nine different systems with platelet recoveries ranging from 17 to 90%, and PDGF-AB release ranging from 33 to 140 ng/mL.[26] By comparison, we found a 98% platelet recovery and an average release of 127 ng PDGF-AB per patch. Calculating the growth factor contents in relation to the number of platelets drawn from the donor has been suggested to enable a better comparison between products. We found 37.1, 27.3, and 0.381 pg per 106 platelets for PDGF-AB, TGF-β, and VEGF, respectively. This corresponds well with growth factors released over 5 hours from Choukrons PRF: PDGF-AB: 27.8, TGF-β: 38.3, and VEGF: 0.55 (recalculated to pg/106).[27]

It is well known that CWF has a strong proteolytic activity and we have previously shown an increased growth factor release from a PRF product under proteolytic conditions.[7] We show that increased PDGF-AB release is seen when the patch is incubated with CWF and that this increase appears to be related with the solubilization of the patch. The continued release of PDGF-AB is in contrast to the instant release seen with thrombin activation of PRP.[22]

To assess the effects of the large numbers of leukocytes included in the patch, we made a direct comparison with simultaneously generated PRP. Comparing PDGF-AB, VEGF, and IL-8 between the two preparations revealed differences as would be expected as platelet and leukocyte concentrations were higher in the patch. The classical PRP method used resulted in an average recovery of 48% platelets into a leukocyte-free PRP. Relating growth factor levels to the platelet numbers in the PRP and patch PDGF-AB release were higher in the PRP. This fits well with the ability to continuously release PDGF-AB from the patch preparation. In contrast, both VEGF (twofold) and IL-8 (70-fold) were increased despite this correction. To relate the levels to clinical application, we calculated the actual levels applied at a certain area. These numbers show approximately threefold, eightfold, and 320-fold increase for PDGF-AB, VEGF, and IL-8, respectively. The highly increased level of IL-8 fits well with the high leukocyte contents.[28] IL-8s are involved in the needed acute response initiating wound healing[29] and have been suggested for the treatment of chronic wounds.[30]

By comparing PRP and patch extracts by antibody arrays, the levels of 10 proteins were found to differ; of these, eight were present at higher levels in the patch extracts. These included growth factors involved in angiogenesis (VEGF and FGF-2) as well as reepithelialization (HGF).[31] Cytokines (IL-8, IL-16, CD40L) and immune modulators (IL-1ra), as well as the soluble pattern recognition molecule, PTX3, which are involved in the body's first line of defense as part of the innate immune response.[32] Surprisingly, the collagenase MMP-8 was present in higher levels in the PRP preparation despite its known production by neutrophils.[33] Finally, leptin was found at higher levels in PRP; leptin has been shown to be involved in keratinocyte proliferation.[34]

In order to assess the biological potential of the growth factors released from the patch, we followed the growth of primary human fibroblasts in response to patch-derived substances. An insert-based assay was set up to prevent the detection of patch-derived cells. The results showed a strong mitogenic effect on the human fibroblasts used, confirming the biological activity of the patch-derived growth factors. We have previously shown a similar effect of both a traditional PRP and the Vivostat PRF.[7] Increased fibroblast migration seen in initial studies confirm the biological effect of both patch and PRP-derived factors, although the effect of PRP seems to be lower than patch at the highest concentrations tested (Supporting Information Figure S2). As leukocyte-derived fibrocytes have been implicated in the healing process, we investigated if these cells could be derived from the patch. We showed that fibrocytes can be grown from the patch. The theoretical benefit of this finding has to be confirmed by further studies.

Acute wounds are known to heal by going through several steps: hemostasis, inflammation, proliferation, and remodeling.[35] Nonhealing wounds are believed to be halted in the inflammatory phase. Studies on acute wounds show that extremely high numbers of neutrophils are present within hours of wounding, but disappear within days,[36] followed by the presence of monocytes and macrophages that transfer the wound to the proliferative phase of wound healing.[35] Although highly complex, recent studies suggest that the apoptosis of neutrophils[37] and its phagocytosis by monocytes/macrophages leads to a decrease in neutrophil migration to the wound and the transfer of the macrophages into a so-called wound-healing phenotype.38,39 Abrogated healing of macrophage-depleted mice confirms the importance of these cells.[40] Other studies have shown weak perivascular infiltrates in diabetic ulcers, despite the up-regulation of attachment molecules known to be involved in the acute inflammation and wound healing.[41] The resulting low-grade chronic inflammation leads to a lack of an acute-focused immune response, essential for the healing cascade to continue.[42] This is in line with a study showing that accelerated healing in mice was associated with an increased infiltration of leukocytes and fibrocytes.[43]

In this study, we show that a leukocytes and platelet rich patch can provide a way of transferring concentrated cells and signals directly to a surface. On that basis, we suggest that the patch described here could be beneficial for the healing of recalcitrant wounds. To investigate this potential, in a relevant setting, clinical testing in humans is needed. A pilot trial on 15 patients, with 16 lower extremity chronic wounds of varying etiologies, has been performed with a positive outcome[44] and a larger multicenter trial on diabetic foot ulcers is currently being performed.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Hanne Vogensen, Lone Haase, and Anette Walther Mørck are acknowledged for their valuable practical assistance; Derek Hollingsbee for the statistical analyses and excellent input. The described studies are supported by Reapplix Aps and The Danish Agency for Science, Technology and Innovation (Grant 09-070818).

Conflict of Interest: KH, CC, BJ, and TK state no conflict of interest. RL is co-inventor of the Leucopatch™ technology.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
wrr870-sup-0001-FigureS1.doc1363K

Figure S1. Relative levels of proteins extracted from the patch (black columns) and a PRP (white columns). Samples are ordered by fold increase in the patch compared with PRP. Abbreviations: MMP-8, matrix mettalloprotease 8; PTX3, pentraxin-3; CD40L, CD40 ligand; HGF, hepatocyte growth factor; VEGF, vascular endothelial growth factor; FGF-2, fibroblast growth factor 2; IL-1ra, interleukin-1 receptor antagonist; IL-16, interleukin 16; IL-8, interleukin 8.

wrr870-sup-0001-FiguresS2-S3.doc17283K

Figures S2 and S3. Migration assay.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.